US2967018A - Analog computation - Google Patents

Description

Jan. 3, 1961 L. E. FOGARTY ANALOG COMPUTATION Filed Jan. 4, 1957 BIL f i t LAURENCE E. FOGARTY INVENTOR ATTORNEY United States Patent Ce 2,967,018 PatentedJanr3, 1961 ANALOG COMPUTATION Laurence E. Fogarty, Binghamton, N .Y., assignor to General Precision Inc., a corporation of Delaware Filed Jan. 4, 1957, Ser. N0. 632,481

2 Claims. (Cl. 235-183) that every mathematical operation performed in an analog computer be performed with as little error and noise as is possible.

A number of differential equations expressing the operation of various physical systems require computation of the time integral of the product of two or more timevarying quantities, where one of the quantities and its time derivative may be available in the computer as voltages, but where only the time integral of the other quantity may be available, perhaps as a shaft-position. The most common prior art manner of solving the problem has been that of differentiating the,shaft-position sec- 0nd quantity with respect to time, multiplying the two quantities, and then integrating the product with respect to time to provide the desired answer quantity. Due to the limitations of various contemporary electrical analog computer apparatus, a number of undesirable effects attend such a method, and some of these limitations of the prior art will be pointed out below more specifically. The invention, on the other hand, entirely avoids the undesirable step of differentiating the shaft position with respect to time, and also provides the desired output quantity with simpler and more accurate apparatus.

Thus it is a primary object of the present invention to provide improved method and means for providing an analog potential commensurate with the product of two time-varying input potentials.

It is a more specific object of the invention to provide improved method and means for providing an analog potential commensurate with the time integral of the product of two time-variables, where the time derivative of one of the variables is available as an electrical analog and where the time integral of the other of said variables is available as a mechanical analog.

Other objects of the invention will in part be obvious and will in part appear hereinafter.

The invention accordingly comprises the several steps and the relation of one or more of such steps with respect to each of the others, and the apparatus embodying features of construction, combinations of elements and arrangement of parts which are adapted to effect .such steps, all as exemplified in the following detailed disclosure, and the scope of the invention will be indicated in the claims.

For a fuller understanding of the nature and objects of "the invention reference should be had to the following detailed description-taken inconnection with'the accompanying drawings, in which:

Fig. l is.an electricalschematic diagram illustrating common prior art apparatus for computing the time integral of the product of two time-variables; and

Fig. 2 is an electrical schematic diagram illustrating the preferred embodiment of the invention. In both figures, well known analog computer elements are shown in symbol form for sake of clarity, and like numerals refer to similar parts.

Referring now to Fig. 1, there is shown typical prior art apparatus for solving the problem mentioned above. An electrical quantity 50' derived by other apparatus not shown exists at terminal 101. This voltage may vary with time in magnitude and polarity in accordance with a physical or numerical quantity it is intended to represent. For example, the quantity i may represent the acceleration, either angular or translational, of a simulated or actual physical member. In order to determine the velocity 03 of such a member, the electrical quantity at terminal 101 is applied through a scaling resistor R-101 to an integrating means 1-101 shown in block form. Integrator 1-101, which may comprise any one of the many well known analog computer integrators, suchas a velocity servo or Miller integrator, integrates the d: quantity with respect to time to provide an output quantity commensurate with at, the velocity of the member. The a: quantity may be applied to a further integrating means indicated by integrator 1-102 in Fig. 1, to provide an output quantity commensurate with the position of the member. A vast number of analog computers solve second order differential equations in which such successive integrations are made in order to simulate the op eration of physical systems. The dashed line at the right-hand side of theintegrator block indicates that in tegrator 1-102 provides a mechanical output position.

able neither as a voltage nor a shaft position, but where the time integral quantity y is available as a shaft posi-' tion, as illustrated by the provision of shaft 102 in Fig. 1. In order to provide the desired product for either electronic or electro-mechanical integration, the prior art has differentiated the quantity represented by the shaft position and then multiplied the two quantities. For example, the prior art manner of attacking the problem may utilize a resistor-capacitor differentiating circuit such as shown in Fig. l. Potentiometer R-102 has its winding excited by constant potentials from the computer power supply (not shown), and its wiper arm positioned in accordance with the quantity y by the position of available shaft 102. The potential commensurate with the quantity y appearing on the arm of the potentiometer is routed to a conventional resistance-capacitance differentiating circuit comprised of capacitor C-101 and resistance-R403. The output of the difierentiating circuit is shown connected to the input terminal of a conventional analog computer position servo. The input voltage to the servo is amplified by amplifier U-101 and applied to drive a conventional servo motive means SM, which mechanically drives a tachometer generator 'TG and a rebalancing or follow-up potentiometer R-104.

The servo motor SM drives the servo output shaft 103 into a 1] shaft position at shaft 103, multiplication of the x and 17 quantities is easily accomplished, merely by exciting a potentiometer R-105 on shaft 103 with the dc voltage from terminal 101. The xy voltages on the arm of potentiometers R-105 then may be integrated either 5 electronically or electromechanically to provide the desired integral of the product, h d-12 dt.

Potentiometer R-102 will have finite resolution, however, so that the y shaft position will not be converted smoothly into a y voltage on the arm of potentiometer R-102. Instead, as shaft 102 is rotated, the voltage on the arm of potentiometer R-102 will increase or decrease in a series of small but sudden steps. When such a y voltage is differentiated, such as by the RC differentiating circuit shown, the 1] voltage will contain sharp spikes 15 able time that it is inadvisable to differentiate the output voltage from a potentiometer, and hence various other means for providing a derivative of the y shaft position have been used. A very common alternative is the use of a tachometer generator, such as shown at G in Fig.

l, driven by the y shaft. Tachometer generator G provides a much smoother 17 voltage than the potentiometer- RC ditt'erentiator combination shown, so that connection of generator output terminal 108 to the servo input terminal 106 in place of the 17 signal shown applied via capacitor C-101 will, in general, provide improved op- 5 eration. However, tachometer generators have finite inductance and limited sensitivity at low speeds, so that the 17 voltage supplied to the position servo from genorator G is liable to be in error, either statically, dy-

namically, or both. And finally, regardless of how the 40 differentiation of the y shaft position is done, the position servo utilized to convert the 1] voltage into a 1] shaft position inherently has limited frequency response and appreciable velocity lag, so that the instantaneous position of shaft 103 is liable to be inaccurate. the numerous inaccuracies of the prior art methods and apparatus, the absence of improved method and apparatus has required use of the former, often to the considerable detriment of overall computer operation.

Referring now to Fig. 2, there is shown an exemplary embodiment of the present invention, w.'th well known parts shown in symbol form. The input potential at terminal 101 is applied via scaling resistor R-101 to an electronic integrator I-101 shown as comprising a conventional operational amplifier having a capacitor 0-201 55 connected in its feedback circuit. Electronic integrator I-101 integrates the it? voltage with respect to time, providing an :1: output voltage in extremely accurate manner,- since electronic integrators, as is well known, are not plagued by the friction, hysteresis, blacklash and inertia 0 limitations of electromechanical integrators. The a: voltage may be integrated again by means of electronic integrator I-103 if an output voltage commensurate with the variable x is desired. If the quantity x is required as a shaft position esewhere in the computer, any conventional position servo may be connected to the output circuit of integrator I-103 to furnish the x quantity in mechanical form.

The d: quantity output voltage from electronic in egrator I-101 is applied to excite potentiometer R-201, the arm of which is positioned by the y-input shaft 102, thereby deriving a potential commensurate with the product 03y. As shown in Fig. 2, such product potential is applied via summing resistor R-205 to a summing device shown as comprising a conventional analog computer Hence the use of a differentiated potenti- 20 In spite of 45 summing amplifier U-102. The 5 input potential from terminal lill also is applied to excite a further potentiometer, R202, the arm of which also is positioned by the y variable input shaft. The potential on the arm of potentiometer R-202, which potential is commensurae with the quantity 12a), is applied via scaling resistor R203 to a second electronic integrator 1402, which integrator provides an output potential commensurate with the quantity Io yfi/dt- Such output potential is applied via summing resistor R-204 of the summing device. As will now be explained, the output of the summing device is the desired product quantity, and it will be apparent at this point, that the desired product quantity has been obtained without differentiating by the y input quantity.

The differential of a product (uv) may be written as follows:

and

Transposing,

udv=d(uv) vd1( Integrating,

judv=uvjvdu Substituting for u, and also substituting y for v in the above expression,

da: da: -dvy fydu da: d e u then (iiiand if y=v then 2 d1) Hz'dt Substituting further into the above expression:

in; Q dz i fit dt Tli y which written with dot notation for time derivatives provides:

Io fl =yfoy As shown in Fig. 2, the two quantities applied to summing amplifier U-102 are commensurate with the two terms on the right-hand side of the above expression, the ray term being applied via resistor R-205 and f yz'zidt term being applied via resistor R-206. As indicated by the analysis above, the sum of these two potentials provides the desired integral of the product without differem tiation.

A typical application of the invention may now be considered. In an analog computer such as a flight simulator or a flight table it may be desirable to compute simulated altitude h of a simulated aircraft or missile, utilizing the formula:

. h=f V sin 'ydt wherein V=velocity along the fiight path 'y==fiight path elevation angle If an input potential commensurate with cos is 0 applied to terminal 101 in Fig. 2, and a shaft position sin 'r y is applied to terminal 101, an output potential commensurate with horizontal distance will be derived at terminal 209. The cos 'y, sin 'y, 7 and 6 quantities are frequently available in desired form in a number of contemporary computers. Numerous further applications of the invention to other computations will become readi'y apparent to those skilled in the art in light of this disclosure.

While Fig. 2 shows a single embodiment of the invention, those skilled in the art will recognize that various equivalent computer techniques may be substituted. As examples, series summing of voltages may be substituted for the parallel addition arrangement shown, and various forms of potential modifying means may be utilized in lieu of the simple potentiometers shown.

It will thus be seen that the objects set forth above, among those made apparent from the preceding description, are efliciently attained. Since certain changes may be made in carrying out the above method and in the constructions set forth without departing from the scope of the invention, it is intended that all matter contained in the above description or shown in the accompanying drawing shall be interpreted as illustrative and not in a limiting sense.

Having described my invention, what I claim as new and desire to secure by Letters Patent is:

1. Analog computing apparatus for providing an output potential commensurate with the time integral of the product of two independent quantities which vary with time comprising in combination, means for deriving an input potential commensurate with the time rate of change of one of said quantities, integrator means responsive to said input potential for deriving a second potential cornmensurate with said one of said quantities, a mechanical shaft position input device having a position commensurate with the time integral of the second of said two quantities, first potentiometer means operated by said mechanical input device for modifying said input poten tial to provide a third potential, second potentiometer means operated by said mechanical input device for modifying said second potential to provide a fourth potential, means for integrating said third potential with respect to time to provide a fifth potential commensurate with the time integral of said third potential, and means for combining said fourth and fifth potentials to provide said output potential.

2. Electrical analog computing apparatus for providing an output voltage commensurate with the time integral of the product of two independent variables comprising in combination, electronic integrator means responsive to an input potential commensurate with rate of change of a first of said variables for providing a first potential, a pair of potential modifying means operable by a mechanical input device for modifying said input and first potentials in accordance with the time integral of the second of said variables to provide second and third potentials, electronic integrator means for integrating said second potential with respect to time to provide a fourth potential, and electrical summing means responsive to said third and fourth potentials to provide said output potential.

References Cited in the file of this patent Electronic Instruments (Greenwood et al.), 1948, p. 131-134.

Electrontube Circuits (Seely), 1950, p. 165.

Project Cyclone, part 2 (Murphy), 1952, pages 43 and 50.

The University of Connecticut Engineering Experiment Station (Robb), pages 4-7, 1953.

A Palimpsest on the Electric Analog Art (Paynter, Editor), 1955, pages 141-145.

Images